Abstract Disorders of social behavior and communication are increasingly prevalent and pose a substantial burden to society. These disorders often show sex differences in prevalence, expression and severity. One explanation for these differences reflects dysfunction in the sexually different social brain. A particularly relevant neuropeptide system in this respect is the vasopressin (AVP) innervation of the brain, which shows marked sex differences across many species, including humans, and which has been implicated in both aggressive and affiliation behavior. AVP fibers are prominent in most areas of the social brain and many of these originate from cells within the posterior bed nucleus of the stria terminalis (BNST). However, so far no studies have directly targeted these cells to test their role in social behavior. Here we do so using viral vector technology to test the overarching hypothesis that these BNST AVP cells promote male-typical affiliation and that they do so by receiving direct social sensory and arousal/reward-related modulatory input. We will test this hypothesis across two specific aims using a multi-disciplinary approach: (i) does removal of BNST AVP cells eliminate prosocial communication in male, but not female, mice? (ii) does viral-vector tracing of monosynaptic connections to BNST-AVP cells reveal sensory and modulatory inputs? The answer to these novel research questions will uncover a fundamental mechanism by which the brain regulates sexually-differentiated social communication, and will contribute to identifying causes of, and treatments for, disorders of social communication.